Professional data scientist here.
Been getting a LOT of questions from college students and recent grads about the major, careers, and the data analytics, data science, machine learning, and AI fields. Figured it was time to have a CC thread.
Ask me anything.
Do you think a double major in Finance and data science or a minor in data science can be valuable?
Absolutely. A dual major, minor, dual minor, or certificate in a content area is a great idea for anyone in the field. It gives you domain knowledge which is a great asset for a data analyst/scientists.
For Finance, you can work it in either direction.
A degree in Finance plus a dual major or minor in DS makes you a much more employable Financial analyst. Look at any job on indeed, and every single entry level finance job is looking for the âholy trinityâ of SQL, advanced Excel skills, and Tableau or some equivalent business Intelligence & visualization tool like Looker, PowerBI, Qlik,Spotfire, etc. Adding some Python/R coding skills is icing on the cake.
If you go as a DS major with a Finance minor, it gives you advanced understanding of business, finance, and accounting, which helps you solve some of the problems that youâre likely to encounter.
My company added a data team this past two years and moved to Power BI from tableau.
The data scientist and her team - donât check, analyze, or have any knowledge of what it means.
They donât interpret. Just build dashboards.
How common is that - as a data guru is one simply building ? I find it unlikely but thatâs how it is at my global company.
Also at schools that donât have a DS major/minor, what majors or minors would best add depth ? Math ? Stats etc ? Which specific classes in those areas ?
So thatâs unfortunate that all they do is dashboarding but it is not unusual. There are fundamentally, three types of teams that youâll see in organizations: âFull stackâ data scientists, Larger Teams with specialized members, and hybrids.
Full Stack - Historically, data scientists have had the skills to do a bit of everything themselves: Acquiring data, pipelining it cleaning it, then doing the feature engineering modeling and deployment. in some organizations, the preferred role remains a full-stack data scientist that does E2E work. usually coming out of university, youâll know one or two ways to do each step in the workflow and you can get by. Itâs impossible to become expert in all of these skills in a short period of time given the variety of problems and platforms So you can continue to build your skill set over many years.
Specialized -The field has matured and begun to split into a number of very specialized roles. But. It is not âone size fits allâ. A larger team tacking large problems might have business analysts, data analysts, data engineers, analytics engineers, machine learning / NLP / AI engineers, and frontend web developers. A smaller team tackling limited goals might just have a few specialties.
Hybrid - Depending on the complexity of the problems and the nature of the business and technical solutions, teams might just employ data analyst and data scientists alongside a few engineers, and let those generalists do a hybrid of system building and data analysis / interpretation as well as solutioning. This is the way I work most of the time.
One specialist is called an analytics engineer. Their role is to build the pipeline and dashboards to deliver the data for others to gain insight and understanding. So, that team you describe sounds like an analytics engineering team not a data science team. If that delivers value, perhaps thatâs sufficient if someone else does the interpretation. Keep in mind that a lot of organizations have data but just need deliver it to users so - if they can just pipeline it and dashboard it - well that can translate to immediate value. What you describe is an essential element everywhere, but as you hinted, you can do a lot more of course on the analysis and modeling. itâs a shame the analytics team isnât participating but that can be how the work is divided.
Regarding your second question on programs and courses at colleges without data scienceâŠ
You probably need to bring together courses or minors in computer science, stats, and some analytics.
People have crowd sourced some good open source âmastersâ in data science. Google them. They are a good starting point.
For comp science, most minors will be a decent core. Usually there are two intro programming courses, and one course each in data structures, discrete mathematics, and algorithms. What you have to supplement with is a deeper dive into a scripting language - Python and/or R - because the CS minor is usually taught with Java, C++, etc which are not the usual data science tools.
For stats, the minor again is a good pairing with the CS minor. You will get (and want) at least 1-2 intro stats courses, a regression course, a Bayesian stats course, and one that covers exploratory data analysis / analytics. Often linear algebra is recommended- I found it to be a course that I could have done without but itâs often required.
Even if there isnât a DS degree, there will often be one or more data analytics or data science / machine learning courses hiding within various departments: stats, physiology, sociology, polisci, business, science, Econ, etc. All of these fields now (really every field) uses DA/DS/ML/AI so often the department will make sure they can train their students and the tools are basically the same. Humanities might prefer R, Econ might prefer Stata, and others python, though it really depends on the school - but the courses definitely exist. Ideally you want to take a series of 2-3 in basic analytics, data mining, ai, and machine learning. Some can be heavy on theory so be careful - try to find ones with more practical problems solving a projects.
Thanks. Thatâs very in depth.
Whatâs interesting is my company has franchisees that sell our products. In lieu of relationships, weâve gone to data. The issue is - no one uses -.âit me as an employee, the franchisees. As one recently told me, how will any of this help me sell more ?
Thatâs my worry. Management has all these ideas but are detached from reality. It may be my industry only. Data is good of course.
But if you are just building but not interpreting - Iâm guessing thatâs not what most go to college for. It will be interesting to see how it all plays out.
Thanks for all the detail on schooling . As Iâve read the past few years on CC it seems like itâs the flavor major of the day.
Based on the first question, I wonder if kids know what they are getting into ?
Having specialist analytics people building state of the art pipeline and dashboards is a best practice. Having no one look at the data is an absolute worst practice
Thank you. This was very helpful! My senior wants to get a job as a risk analyst/consultant after college and is hoping that a data science minor or double major will be beneficial.
What is the difference between studying data science, data analytics and business analytics?
I would love to say it is clear cut but the degrees are too new and havenât been standardized. And job titles arenât helpful.
The main difference in undergraduate degrees probably stems the most from the depth of math, the extent of coding, and the school/college that awards the degree (computer science, math or stats dept, biz school, social sciences).
A business analytics program might only go as far as calc 1 while a data science program might be heavier. A data analytics program could be either and probably depends on its home school.
The same is likely true of coding. Business analytics in B school generally might be lighter on code and heavier on business intelligence software and visualization while a data science program might be code-first, delve deeper into algorithms, and only touch pre-packaged visualization software occasionally.
But, it will vary. There are often enough flexibility in electives to swing the content either way.
starting with the school is probably wise - programs hosted by CompSci departments feel like CompSci degrees.
Very helpful! For my senior who isnât as interested in comp sci, but definitely likes math and analysis, do you think business analytics would be a better choice than data science?
anything in a comp sci department or a joint program with one will be more focused on code for sure, so perhaps consider one outside of a CS school in business or arts and sciences.
But there are other unique options beyond business schools. there are growing number of social data programs that range from pretty CS heavy (social Data Science at Maryland) to more nuanced programs (SoDa at Penn State). Then There are GIS programs all over that combine cartography, satellite imaging, etc with data science. There are programs in classic humanities like psych, political science, and economics as well. The econometrics course which is common to most (all?) Econ majors is basically a course in regression which is a core data analytics/science skill.
But again, YMMV, you have to look at the courses to see if it leans more towards analytics (on paper) or data science.
So what role/kind of job would someone with a Business Analytics degree from a business school typically do?
Does it make a difference it is paired with a Finance or Accounting degree?
Are they really two different career path or potentially complementary?
A typical entry-level financial analyst job right now, would commonly ask for a degree in some business field, like finance or accounting, and ask for analytical skills so the analytics minor, or dual major would provide those skills If youâre going after a job in finance or accounting or auditing, whatever.
If you wanna get a business analytics degree, and work in a field thatâs not necessarily just âbusinessâ it will depend on the company. Some companies like mine would be very comfortable hiring a business analytics degree into a variety of business or more technical jobs where they itâs it just matters that you have the analytic skill set. There are other companies that would want to see analytics paired with some other specific domain knowledge like HR, supply chain, etc. Unless you had some other experience in the industry. Some companies are just picky about what they want.
What I would recommend is, if youâre going to get a business analytics degree that doesnât have a specific business concentration, like supply, chain, HR, MIS, accounting, finance, entrepreneurship, etc. leading to an obvious job then do at least a minor or dual major in some unrelated field to complement the analytics degree and that way you could go two different ways, depending what jobs Come up. Likely itâll matter for your first job because once you start working as a practicing data professional, thatâs gonna matter more down the line
Two examples : one preferring a specific degree and one more general where business analytics or really any quantitative degree would be fine
First one is in Chicago for Kraft Heinz and is one level above a college hire but something like this should exist for college recruiting. the requirements are broad but some kind of specific business or Econ degree is preferred.
Example 2. Atlanta. Kimberly-Clark paper company
No specific degree required - just the usual analytics tools : Tableau and other BI tools, excel skills, usually SQL (though not in this specific posting). You could be business analytics or you could be history or psych with an analytics minor provided you had the key skills and maybe an internship
